One of the most difficult challenges in mobile robotics is real-world navigation. A real world can change suddenly and this change makes the robot relinquish planning actions in advance. In order to overcome such a change, behavior-based navigation was introduced. However, it had a difficulty in planning deliberate actions and in communicating with humans.

We propose a new control strategy combining both the merits of behavior-based and planner-based approaches. The architecture consists of three major parts: Behaviors, Planner, and Coordinator. The Planner plays two important roles: 1) as a flexible human interface and 2) as the planner itself. The Coordinator serves as an interface between Behaviors and Planner and guides Behaviors to accomplish meaningful tasks according to the guidelines from the Planner and the Position estimator.

We also provide a brief description of the intelligent mobile robot CAIR-2 and for '95 IJCAI/AAAI Robot Competition and Exhibition when the robot was placed first in the Office Delivery event.